Semester

Fall

Date of Graduation

2023

Document Type

Thesis

Degree Type

MS

College

Statler College of Engineering and Mineral Resources

Department

Mechanical and Aerospace Engineering

Committee Chair

Nicholas Szorcinski

Committee Member

Xi Yu

Committee Member

Sergiy Yakovenko

Committee Member

Yong-Lak Park

Abstract

Researchers have been studying the mechanisms underlying animal motor control for many years using computational models and biomimetic robots. Since testing some theories in animals can be challenging, this approach can enable unique contributions to the field. An example of a system that benefits from this modeling and robotics approach is the campaniform sensillum (CS), a kind of sensory organ used to detect the loads exerted on an insect's legs. The CS on the leg are found in groups on high-stress areas of the exoskeleton and have a major influence on the adaptation of walking behavior. The challenge for studying these sensors is recording CS output from freely walking insects, which would show what the sensors detect during behavior. To address this difficulty, 3 dynamically scaled robotic models of the middle leg of the stick insect Carausius morosus (C. morosus) and the fly Drosophila melanogaster (D. melanogaster) were constructed. Two of the robotic legs model the C. morosus and are scaled to a stick insect at a ratio of 15:1 and 25:1. The robotic fly leg is scaled 400:1 to the leg of the D. melanogaster. Strain gauges are affixed to locations and orientations that are analogous to those of major CS groups. The legs were attached to a linear guide to simulate weight and they stepped on a treadmill to mimic walking.

Using these robotic models, it is possible to shed light on how the nervous system of insects detects load feedback, examine the effect of different tarsi designs on load feedback, and compare the CS measurement capabilities of different insects. As mentioned earlier, robotic legs allow for any experiment to be conducted, and strain data can still be recorded, unlike animals. I subjected the 15:1 stick leg to a range of stepping conditions, including various static loading, transient loading, and leg slipping. I then processed the strain data through a previously published dynamic computational model of CS discharge. This demonstrated that the CS signal can robustly signal increasing forces at the beginning of the stance phase and decreasing forces at the end of the stance phase or when the foot slips.

The same model leg can then be further expanded upon, allowing us to test how different tarsus designs affect load feedback. To isolate various morphological effects, these tarsi were developed with differing degrees of compliance, passive grip, and biomimetic structure. These experiments demonstrated that the tarsus plays a distinct role in loading the leg because of the various effects each design had on the strain. In the final experiment, two morphologically distinct insects with homologous CS groups were compared. The 400:1 robotic fly middle leg and the 25:1 robotic stick insect middle leg were used for these tests. The measured strains were notably influenced by the leg morphology, stepping kinematics, and sensor locations. Additionally, the sensor locations were lacking in one species in comparison to the other measured strains that were already being measured by the present sensors. These findings contributed to the understanding of load sensing in animal locomotion, effects of tarsal morphology, and sensory organ morphology in motor control.

Download

Share

COinS